Helium-Uranium Ratios for Pleistocene and Tertiary Fossil Aragonites

Helium-Uranium Ratios for Pleistocene and

Tertiary Fossil Aragonites

Abstract. Uranium:helium ages are compared with Th230: U23' ages and stratigraphic ages for selected Pleistocene and Tertiary shells and corals. Agree- ment is good except for those fossils for which independent criteria (for example, aragonite :calcite, Ra226: Th230, or U234: U238 ratios) indicate a chemically open system. Certain of the Tertiary samples yield erroneously low ages which probably result from secondary addition of uranium. The helium geochronometer appears to yield reliable minimum ages and may be applicable to direct dating of aragonitic fossils-especially those of Pleistocene age-when used in conjulnction with other isotopic and geologic information.

Several proposed correlations of the absolute and paleontologic time scales have been reviewed by Kulp (1). Dur-

ing the last few years absolute ages assigned to most critical points in the paleontologic time scale have changed very little and are generally agreed upon. Hence the mere presence of a

reasonably complete Paleozoic or Mes- ozoic faunal assemblage normally per- mits more precise absolute chronology of a site than the direct application of absolute dating methods of phases oc- curring at the site. However, in the late Tertiary and Pleistocene it is fre-

quently impossible to derive the absolute age with satisfactory precision from the faunal assemblage alone ow-

ing to the short time-spans involved and the slowness of evolutionary change. Moreover, it is in just this time range that the dating of associated igneous rocks is frequently very difficult, primarily because the very short absolute time spans involved seldom include enough separate igneous events at a specified site to provide satisfactory time resolution. Thus, development of a geochronological method directly applicable to fossils would be valuable for problems of absolute Pleistocene

geochronology, and would almost cer-

tainly yield further information con- cerning Pleistocene eustatic changes, glacial cycles, human chronology, and related problems.

From an analytical point of view, the helium method has advantages over the argon method: the production rate of helium in pure carbonate fossils

greatly exceeds the argon-production rate, and the problem of atmospheric contamination is much less serious because helium abundance in the atmo- sphere is only 0.05 percent of argon abundance. In order that the helium method may be useful, however, certain additional conditions must be approx- imated.

312

If uranium was incorporated in a shell or coral during or shortly after the life of the animal and the fossil did not subsequently exchange uranium or radioactive daughters with its surroundings, then the helium-production rate at any time in the past can be calculated from knowledge of the present concentrations of uranium isotopes. If helium now present in the fossil resulted only from decay of uranium and its daughters and no helium was lost to the surroundings, then measurement of present concentrations of uranium and helium yields the age of the fossil. Certain other initial conditions, such as absence of initial Th230, must also be known; these will be discussed in connection with the method of age calculation.

Studies of the helium method (2) have shown that low ages based on helium, obtained on common rock- forming minerals, do not necessarily reflect diffusive loss of helium from the lattices of those minerals; under ideal conditions, some mineral lattices even appear to retain helium quantitatively for longer than 108 years.

The distribution and history of alpha emitters in the sample may be far' more critical parameters. It is known, for example, that living mollusk shells contain far less uranium on the average than do similar fossil shells, in-

dicating that most of the uranium in the latter is probably secondary (3). However, fossil shells in the 104-year range of age have concentrations of uranium no less on the average than those of fossils in the 106-year range; it seems that the process of net uranium addition may cease a few thousand years after the death of the animal. However, the possibility of uranium exchange between the fossil and its surroundings is not eliminated by such data; in some instances exchange could be detected by measurement of

U234:U238 ratios in the fossils (see 3). We have attempted to determine the

effects of possible helium loss and uranium exchange on the helium ages of aragonitic fossils by performing helium- age determinations on various samples of fossil shells and coral for which several types of independent estimates of age are available. Specifically, we have made three types of comparison to test the above assumptions: (i) we compared the He:U ratios of several fossils known to be of the same age but having very different uranium contents; (ii) we compared helium ages with ionium ages; and (iii) we compared helium ages with well-defined stratigraphic ages.

The helium contents of these fossils were determined by use of the statical- ly operated, gas mass spectrometer (4). Aragonite samples (< 1 g) were de- crepitated in vacuo by means of radio- frequency induction heating. Evolved CO, was frozen on a cold trap at liquid-nitrogen temperatures; the trap was then isolated and the gas remaining in the rest of the system was gettered with hot Ti. Finally, other cold and charcoal traps were immersed in liquid nitrogen to remove any gases remaining after chemical clean-up. The total pressure of gas remaining after this operation, of which helium was a minor constituent, was usually in the neighborhood of 10-" mm-Hg; when this pressure was achieved the helium was introduced into the tube. The spectrometer was repeatedly calibrated with a gas pipette that admitted a known amount of He4 to the tube; sensitivity was constant within about 10 percent for months. The entire vacuum line, except the spectrometer tube, was constructed of 1720 Pyrex glass to minimize inward diffusion of atmospheric helium.

The most serious analytical limita- tion encountered was that modern samples, especially of coral, which should have contained virtually no helium, often yielded apparent helium contents [3 to 6 X 10-9 cm3 total He at standard temperature and pressure (STP)] exceeding that of the nor- mal "empty crucible" system blank (- 1 to 3 X 109 cm3 at STP). We have not yet made sufficient measurements to define the origin or constancy of this "nonradiogenic" helium; the possibility that it represents the decay of great initial excesses of relatively short- lived intermediate daughters (for ex-

SCIENCE, VOL. 149

ample, Ra226) can be discounted, since sufficiently great excesses of this type have not been observed in corals or shells of any age (3). This problem does not seriously affect interpretation of most ages presented in this paper, but if an effort were made to overlap the C14 range with the helium method, the origin and variability of this finite "sample blank" would have to be thoroughly understood. If it is related to the presence of very small amounts of detrital material in the sample, sub- stitution of acid solution for the de- crepitation procedure may eliminate the problem.

Routine fluorometric measurements of uranium contents at Lamont Geo- logical Observatory (3) were checked, on selected samples, by isotope-dilution techniques (alpha spectrometry at La- mont and mass spectrometry at Brook- haven National Laboratory). The methods employed in the alpha spectromet- ric measurement of contents of U238, U234, and Th230 in shells and corals and the calculation of the corresponding ages are discussed by Kaufman and Broecker (5) and Thurber et al. (6).

Reproducibility of both uranium and helium measurements was normally within 10 percent. When isotope-dilution results and fluorometric results were compared on different portions of the same samples, they agreed within this error, as did helium replicates measured by the two spectrometers, in which sensitivities and amounts of He4 admitted by the spike pipettes both differed by a factor exceeding 3. A 15-percent analytical error has therefore been assigned to the ages. In some instances 'the apparent "sample blank," already discussed, was a significant factor in increasing this uncertainty; a result was that in certain instances ages could not be assigned.

The first test of the assumptions of the method was to obtain U:He ratios for shells of identical age but of different uranium content. Each of several raised beaches in California provides various shells of identical age (5). One such suite studied was from terraces that correlate with the 70- foot marine terrace of the Palos Verdes hills; the other was from another Palos Verdes terrace 1230 feet (375 m) above the present sea level. Results on these samples appear in Fig. 1 and Table 1; the latter includes results on

16 JULY 1965

30.

!- c3

gO

I 0

0> %ci)

20

10

0 L .r . I ..,.I -.1 I

.......

I I....

0 1 2 3 4 5 6 7 8

U (ppm)

Fig. 1. Helium content of selected mollusk shells as a function of uranium content. Note that within each suite the He:U ratio is essentially constant despite great variations in uranium content. Note in particular the point almost coincident with the origin. The two suites fall on isochrons having quite different slopes as expected on the basis of the Th230:U'24 data (Table 1) and on geologic grounds.

E c 5.0

I- (n t1 E 3.0 I 0

2.0

2 3 Time (105 years)

Fig. 2. Helium:uranium ratio versus time for closed systems having different initial U23: U23s ratios. Note how the lines flex upward as Th'2' and its daughters (assumed to be initially absent) build into transient equilibrium with the U28'. Note also that (to the right of the area shown) the lines must become parallel as secular equilibrium is approached, that is, as unsupported U234 disappears.

313

a gastropod sample from the Lomita cannot be completely eliminated. How- marl which is thought, on geological ever, such data, together with the gen- grounds, to be intermediate in age be- eral agreement between C14 and tween the two terraces. The field rela- Th230:U234 ages on shells and the tionships of all these samples are dis- evidence from concentrations of ura- cussed by Kaufman and Broecker (5). nium discussed earlier, tends to indi- The data in Table 1 are not cor- cate that later episodic or continuous rected for possible "nonradiogenic" or addition of uranium is not the general "initial" helium; as we show later, this rule. The Palos Verdes suites were component of total helium must be selected for Th230:U234 dating by negligible for these suites. Despite the Broecker and Kaufman largely on the wide range of uranium contents in the basis of their aragonite calcite and results, consistency of the He:U ratio Th230: Ra226 ratios (5). within each suite is within the limits A conclusion to be drawn from Fig. of experimental error. This may indi- 1 is that, at least in these shells, the cate that the uranium was added dur- helium observed was produced solely ing the early history of the shells; by radioactive decay of uranium and if it was added during times compar- its daughters in the shells. The fact able with the age of the shells, that the helium correlates quantitative- then different shells having different ly with the uraniutm within each suite propensities to accumulate uranium indicates that (i) no significant amounts (and having, therefore, different ura- of helium were initially present in nium contents) would not necessarily these shells, and (ii) that no significant be expected to have identical He:U amount of the total yield of He from ratios. On the basis of such data alone, the purified samples can be attributed the possibility that addition of uranium to extremely small amounts of extreme- was episodic some time significantly ly old detritus that may still be as- later than the death of the animal sociated with them. This type of

Table 1. Comparison of He:U ages with Th2'0:Un24 ages (14) [< million years (my)] and stratigraphic ages (> 1 my) for various mollusk shells, and the shell contents of helium and uranium. (P), pelecypod; (G), gastropod; (STP), standard temperature and pressure.

U He, STP Age, Independent (ppm) (10-8 cm3/g) U:He estimate of age

1200-foot terrace, Palos Verdes hills, California (G) 807E 1.5 7.7 390,000 ?- 60,000'| 5 samples, > 300,000 (P) 807A 1.7 9.2 420,000 ?- 60,000 1 sample, 325,000 ? (G) 807D 7.0 29. 330,000 ?- 50,000 60,000

70-foot terrace, Palos Verdes hills (P) 853A 0.1 < 0.2 < 200,000 (P) YC-1 .9 .74 105,000 ? 30,000 (P) 853C 1.2 1.3 130,000 ? 20,000 Range (P) SS-1 2.0 1.9 115,000 ? 20,000 120,000 to 140,000 (P) RS-1 2.1 2.0 1.15,000 ? 20,000 (> 20 samples) (G) GS-1 3.8 2.7 95,000 ? 15,000 (P) 855D 4.2 4.7 130,000 4- 20,000

Lomita marl 829 DF 2.4 3.4 155,000 - 30,000 185,000 ?- 20,000

(3 samples)

Israel (G) MIS-1 0.36 Collected alive (G) IS-20 1.8 .76 20,000 ? 2,000 (G) IS-1 3.1 15.3 290,000 ? 50,000 360,000 ? 60,000

Lake Lahonton, Nevada (P) 62-41 0.33 Modern (G) 62-3 .79 1,200 (CT")

(G) 62-21 2.4 17. 440,000 4 60,000 400,000 + 200,000 -- 100,000

(G) 62-22 3.2 25. 500,000 ?- 60,000 Same bed as 62-21

Plum Point, Maryland (P) MCL-1 0.88 240 23 ? 4 my Upper Oligocene to Lower (P) MIF-1P .58 250 36 ? 6 my Miocene, 25 to 30 my

Paris Basin, France (G) EA-1 13 340 2.4 my) Eocene, 50 my (G) EA-2 17 390 2.3 my

E

(acid leached)

314

"isochron" check is strictly valid only for the particular suites of samples studied, but the general validity of the assumed, quantitative, cause-effect rela- tion between alpha decay in Pleisto- cene shells and their helium content is also confirmed by the very small amounts of helium observed in modern and C14-range shells (see Table 1), and by the fact that helium ages did not significantly exceed independent estimates of age for any of the fossils studied (Tables 1 and 2). These observations indicate that helium ages provide reliable minimum estimates of age for this type of sample.

We shall now consider the second test of the assumptions: comparison between helium ages and independent T'h2 80;U23: 4 estimates of age, for which the helium ages were calculated in the following manner.

The helium age of a fossil, a few million years old, can be. accurately calculated from its He:U ratio if one assumes that the helium production rate was constant and equal to the alpha-flux resulting from the uranium content of the fossil, together with all alpha-emitting daughters in secular equilibrium. However if the age is less than about 10 years it is necessary to take into account the initial radioactive disequilibrium conditions in the shell and the rate at which secular equilibrium is achieved. For the ages given here, we have assumed that both Th230 and Th232 concentrations were initially zero in these samples (5, 6) and that the present U2a4 : U238 ratio represents the initial ratio modified only by decay of unsupported U234

during the history of the sample. The age can be calculated by summing the integrated helium production rates, from t,, to t, of what are considered, for convenience, to be three separate helium sources: (i) the excess U234 that decays from its initial value to the supported concentration, (ii) the contribution from Th230 and its daughters as they augment into transient equilibrium with the U234, and (iii) the contribution from the (constant) concentration of U2as. The contribution by U235 and its alpha-emitting daughters adds a minor increment to the total results.

The total accumulated production of helium for any given uranium content is plotted as a function of time and of the initial U234:U238 ratio in Fig. 2; it can be seen that the rate of production of helium increases slowly

SCIENCE, VOL. 149

with Th230 build-in from to (death of the animal), when production of helium results entirely from uranium- isotope decay, until t > 105 years, when most helium is produced by decay of Th230 and its daughters. Note that for young samples the age corresponding to any given helium:uranium ratio can be a significant function of the initial activity ratio of U234:U238. The line corresponding to the present ocean-water value (1.15) is indicated, and most freshwater bodies studied now show values < 2.0 (7). Eventually, as t exceeds 106 years (not shown in Fig. 2), and the excess U234 disappears, the lines become parallel, and the separation between them (which remains constant) becomes insignificant relative to the increasing value of the He:U ratio. Thus the age becomes, for Tertiary samples, a function purely of the observed helium-uranium ratio. For all calculations of age reported by us, the effect of the present U234:U238 ratio (as measured on these samples by Broecker et al.) was taken into account.

In Table 1 Th930: U934 ages obtained on selected mollusk shells by Broecker et al. are compared with the estimates of age based on helium; the shells selected show good agreement between the two estimates in most instances. Note that in the case of the samples from Lake Lahonton and Israel the C14-range and modern samples yielded less than 5 percent of the helium yielded by samples from the same sites for which the age comparisons are made, indicating that the possible significance of initial helium in the latter can be discounted. Very young samples from the Los Angeles County sites were not run, but, as stated earlier, the same conclusion in these instances derives from the fact that the points in Fig. 1 fall close to helium isochrons drawn to pass through the origin; ages based on helium are compared with the range of Th230: U234 ages found for several other shells from the same terraces.

We have made a similar comparison of age for some aragonitic coral samples. A number of cores up to 4000-feet long and extending well into the Tertiary exist for certain coral reefs in the Pacific; their upper parts have been dated by the ionium method (8, 6) and certain of them are well correlated with the paleontological time scale in the Mid- and Lower-Miocene (9). Helium and uranium contents along 16 JULY 1965

Table 2. He ages compared with Th230 : U23t ages for samples of coral, together with their uranium and helium contents. Since recent corals repeatedly yield helium equivalent to *-40,000 years of helium production, it should be borne in mind that helium ages of the older samples may be similarly affected by this violation of the initial conditions assumed in the method; my, million years; STP, standard temperature and pressure.

Sample U He, STP Age, Age, (feet) (ppm) (10-8 cm3) U:He Th230 U234 (feet)

Eniwetok < 25 3 1.3 ? 0.3 40,000 + 10,000 < 5,000 yr

(Av. of 3) (Av. of 3) (Av., equivalent) 50 3.0 3.8 155,000 + 30,000 115,000 ? 5,000 85 2.6 5.4 220,000 ? 50,000 240,000 ? 60,000

100 A 2.3 9.0 380,000 6000 300,000 +

140000 - 40,000 100 B 2.3 9.1 380,000 ? 60,000 225,000 ? 40,000* 140 4.5 18.6 400,000 ? 60,000 235,000 ? 30,000: 170-190 3.1 12.6 400,000 ? 60,000 280,000 ? 60,000 700 1.51 138 8.5 + 1.5 my Near Miocene-Pliocene

boundary, - 11 my 2000 0.70 217 26 ? 4 my Lower Miocene, -23 my

Bikini 900 2.5 147 5.9 ? 1.0 my Mid-Miocene, --17 my

* "Initial" U23: U238 =A 1.15 - 0.04 (6).

with ages based on helium of several samples from these cores are listed in Table 2. The independent estimates of age are based on the Th230:U234 ratios or stratigraphic positions of the cores. Of particular importance is the initial U234:U238 ratio, which is calculated from the present ratio and the ionium age of the sample.

The present U234:U238 ratio in the oceans is about 1.15, and there is con- siderable evidence that this has not changed drastically during the last few hundred thousand years; for this rea- son Thurber and Broecker (6) consider that, if initial U234: U238 ratios in coral samples are not 1.15 ? 0.03, the Th230:U234 ages of the samples are suspect. Note that in Table 2 the three samples showing the greatest dis- crepancy between ages based on helium and Th230:U234 ratio are those whose initial U234:U238 ratios do not fall within Broecker's acceptable range, while those for which the two age estimates are essentially in agreement exhibit ratios that overlap this range within the limits of experimental error. Note also that in the anomalous instances the ionium ages show a reversal with depth, which is not expected on any stratigraphic basis; in no case does the age based on helium exhibit such a reversal. On the other hand, any exchange of uranium isotopes is probably accompanied by an effect on the helium age.

It can be concluded on these bases that, as with the shells, there is no evidence of significant loss of helium

by Pleistocene coral samples. In the discordant instances either method or both may be in error owing to uranium-isotope exchange. One may pre- dict that the ionium ages would be strongly affected in such instances since, for samples older than 200,000 years, the Th230: U234 and U234: U238

ratios are such insensitive functions of time. However, the general consistency of the two methods in the remaining instances, together with the continual increase in apparent age with depth, increases the validity of the results by both methods on these samples.

We have also measured the helium ages of some Tertiary corals from the same cores; the samples are known to be Upper, Middle, and Lower Mio- cene in age, and are unusual for such old coral in that they have not been extensively recrystallized to calcite. Again the agreement in age between the helium results and the estimates based on stratigraphic position is satisfactory for the Eniwetok samples. One of these samples appears to have retained helium quantitatively for up to 30 times longer than the history of the oldest Pleistocene sample; on the other hand it is disturbing that the Bikini coral sample, which exhibits a typical uranium content for unaltered coral (- 3 parts per million), yields the anomalous age, while samples having low contents of uranium yield concordant ages.

Ages based on helium were also determined for some Tertiary Pelecypod shells from near Plum Point, Maryland

315

(Table 1), which are thought to be of lowermost Miocene or Oligocene age. The ages are in rough agreement with estimates based on stratigraphy, and there is in any event no suggestion of extensive loss of helium.

Helium ages of an Eocene gastropod also are compared (Table 1) with a

stratigraphic estimate of age; there is no agreement, and yet the stratigraphic estimate is known to be highly reliable. It is difficult to explain this extreme

anomaly in view of the agreement of

ages based on helium with independent estimates of age for younger shells and corals. The possibility that surface mineralization by uranium may be re-

sponsible for the low helium age can be discounted; essentially the same data came from a similar but acid-leached specimen.

Both concordant (10) and highly discordant (11) helium ages have been previously reported for Paleozoic and Mesozoic carbonates. The observation of both essentially concordant and

highly discordant helium ages despite the limited sampling is therefore not surprising.

In summary, helium content has correlated quantitatively with uranium content for suites of aragonitic shells of equal age. In addition, helium contents of C14-range and recent shells were found to be typically negligible relative to helium contents of com- parable shells from the same localities that were 1 0 to 107 years old. In

virtually all instances where reliable "control" ages were available, ages based on helium proved equal to or lower than "control" ages, whether these were established by Th230:U234

dating (in the < 3 X 105 year range) or by stratigraphy (> 1 million years).

On the basis of these observations we conclude that, in the absence of

recrystallization, the helium content of

young aragonites is typically equal to the total alpha-flux in them since their formation. This means that no significant contribution to the total helium content normally results from incor- poration of primary helium at the time of formation of biogenic carbonate. It also means that contribution of helium from old detritus can easily be reduced to negligible levels for samples such as we examined. Thus it appears that determinations of age from helium, such as we describe, can provide reasonably reliable minimum estimates of age for aragonitic samples.

316

These observations are not surprising in view of the fact that the two most probable violations of the assumptions of the method (secondary addition of uranium and diffusive loss of helium) would both tend to lower helium ages from the true values. For both shell and coral, the greatest nega- tive deviations of ages based on helium from true ages appear in the older

samples (> 107 years), suggesting that diffusive loss of helium may have reached significant proportions for these. In any event, the general agreement between helium and Th23"): U234 ages for the Pleistocene shells and corals indicates that diffusive loss of helium was probably negligible in such instances. We should point out that some of the older coral and shell

samples also yielded reasonable helium

ages despite the fact that they were up to 100 times older than the Pleistocene

samples for which helium and ionium

ages were compared. Secondary addition of uranium ap-

pears to be the most serious potential problem in the method, especially for Pleistocene samples. It is known that

many old mollusk shells exhibit U234: U238 activity ratios very much greater than 1.00, even though they are tens of U234 half-lives old and should therefore show U234: U238 activity ratios of 1.00 within experimental error (12). This means that even though net addition of uranium may have ceased, a significant amount of isotopic exchange of uranium with the environment must have gone on in such cases

during the last million years. It is obvious from electron pho-

tomicrographs (13) that domains of intact aragonite in mollusk shells do not normally extend for distances exceeding a few microns, and that interstitial organic domains are less than 0.1 , thick. Since the alpha range in solids is tens of microns, it might be

expected that helium distribution or retention in such a structure would not be dependent on whether alpha emitters were located in the aragonitic "bricks" or in the extremely thin bands of interstitial organic material between the bricks. However, if the alpha emitters are located in such interstitial sites and if, as a result, a significant amount of exchange of uranium takes place between shell and environment, the .helium-production rate at any time in the past can no longer be accurately estimated from the measured content of

U238. In such an instance, even though the helium content of the shell might still accurately represent the integrated alpha flux, the age could not be accurately determined from the measured contents of helium and uranium. It is also probable that the Th230: U234 age of the shell may be likewise affected by such addition or exchange of uranium; if all uranium were added in an episode late in the history of the fossil, the helium and Th230:U234 methods might yield the same age (or ,t for the system). Yet this age might not represent the time of formation of carbonate. Since no absolute estimates of age are available for these Pleisto- cene samples other than the Th230: U234 ages, it is difficult to estimate the seriousness of the above problem. Virtually the only other means of esti-

mating the absolute age of such young samples would be the K-Ar method, results of which might be compared with U:He ages in a few exceptional Pleistocene localities where fossil aragonite and appropriate igneous phases could be stratigraphically related.

It may be possible to evaluate the seriousness of this effect from the con-

sistency of He:U ratios or ages for several samples (in a suite) having vast-

ly different uranium contents, as well as on the basis of agreement with ionium ages.

In conclusion, it appears that He-U

ages in shells and corals are essentially equivalent to ionium ages of these ma- terials except for the greater range of the helium method. It is unlikely that an isolated age on a given shell would yield a reliable estimate of the fossil age. However, within the limita- tions imposed by our sampling, it appears that the helium method may be of use in Pleistocene geochronology when suites of aragonitic fossils are obtained and when other types of geologic or isotopic information are available.

F. P. FANALE O. A. SCHAEFFER

Department of Chemistry, Brookhaven National Laboratory, Upton, New York 11973

References and Notes

1. J. L. Kulp, Science 133, 1105 (1961). 2. P. M. Hurley, Nuclear Geology, H. Faul,

Ed. (Wiley, New York, 1954), p. 301; P. E. Damon and W. D. Green, Proc. IAEA ICSU Symp. Radioactive Dating Athens, November 1962; F. P. Fanale and J. L. Kulp, Econ. Geol. 57, 735 (1962).

3. W. S. Broecker, J. Geophys. Res. 68, 2817 (1963).

SCIENCE, VOL. 149

4. 0. A. Schaeffer, Report BNL-581 (Brook- haven National Laboratory, 1959).

5. A. Kaufman and W. S. Broecker, in preparation.

6. D. L. Thurber, W. S. Broecker, H. A. Potratz, R. L. Blanchard, Science 149, 55 (1965).

7. D. L. Thurber, Proc. IAEA Symp. Radio- active Dating Athens, November 1962.

8. W. H. Sackett and H. A. Potratz, in Schlanger et al., USGS Prof. Paper 260-BB (1963).

9. K. O. Emery, T. I. Tracey, H. S. Ladd, USGS Prof. Paper 260-A (1954).

10. E. Burksen, N. Kapustin, I. Potupvo, J. Ukr. Acad. Sci. 2, 47 (1934).

11. F. P. Fanale and J. L. Kulp, Amer. Min- eralogist 46, 155 (1961).

12. D. L. Thurber, personal communication. 13. C. Gr6goire, J. Biophys. Biochem. Cytol. 3,

797 (1957). 14. Determined by Broecker, Kaufman, and

Thurber at Lamont Geological Observatory. 15. We thank W. S. Broecker, D. L. Thurber,

and A. Kaufman for helpful discussions during the preparation of this paper and for supplying many of the samples. We also acknowledge the advice and cooperation of H. S. Ladd and J. I. Tracy, Jr., USGS, who supplied most of the coral samples. R. L. Batten, American Museum of Natural History, supplied some of the pre-Pleistocene mollusk shells. We also thank P. H. Abel- son and P. E. Hare, Carnegie Institution, and J. Owen, USGS, who helped us collect Miocene shells from Chesapeake Bay. Work sponsored by the AEC.

19 April 1965

Carbon Abundances in

Chondritic Meteorites

Abstract. Combustion analyses of total carbon in chondritic meteorites indicate a fractionation of this element between the various types of chondrites. The median values for the percentage of carbon by weight are 0.40 for enstatite chondrites, 0.09 for olivine-bronzite chondrites, and 0.08 for olivine-hypersthene chondrites. Olivine-pigeonite chondrites show great variations in their carbon contents.

The abundances of carbon have been determined in 18 chondritic meteorites by a combustion technique. Powdered samples ranging from 0.25 to I g in weight were burned in an oxygen at- mosphere in a LECO induction furnace. One gram of iron chips and 1 g of copper chips were added to acceler- ate the reaction. The combustion prod- ucts were carried in the oxygen stream through a MnO2 trap to remove any SO2 formed, and through a catalyst furnace to convert any carbon monox- ide to carbon dioxide. The total CO2

produced was determined by a modified Orsat technique (see 1) in a LECO model 572-100 carbon analyzer, by dis- solving the CO2 from the oxygen car-

4. 0. A. Schaeffer, Report BNL-581 (Brook- haven National Laboratory, 1959).

5. A. Kaufman and W. S. Broecker, in preparation.

6. D. L. Thurber, W. S. Broecker, H. A. Potratz, R. L. Blanchard, Science 149, 55 (1965).

7. D. L. Thurber, Proc. IAEA Symp. Radio- active Dating Athens, November 1962.

8. W. H. Sackett and H. A. Potratz, in Schlanger et al., USGS Prof. Paper 260-BB (1963).

9. K. O. Emery, T. I. Tracey, H. S. Ladd, USGS Prof. Paper 260-A (1954).

10. E. Burksen, N. Kapustin, I. Potupvo, J. Ukr. Acad. Sci. 2, 47 (1934).

11. F. P. Fanale and J. L. Kulp, Amer. Min- eralogist 46, 155 (1961).

12. D. L. Thurber, personal communication. 13. C. Gr6goire, J. Biophys. Biochem. Cytol. 3,

797 (1957). 14. Determined by Broecker, Kaufman, and

Thurber at Lamont Geological Observatory. 15. We thank W. S. Broecker, D. L. Thurber,

and A. Kaufman for helpful discussions during the preparation of this paper and for supplying many of the samples. We also acknowledge the advice and cooperation of H. S. Ladd and J. I. Tracy, Jr., USGS, who supplied most of the coral samples. R. L. Batten, American Museum of Natural History, supplied some of the pre-Pleistocene mollusk shells. We also thank P. H. Abel- son and P. E. Hare, Carnegie Institution, and J. Owen, USGS, who helped us collect Miocene shells from Chesapeake Bay. Work sponsored by the AEC.

19 April 1965

Carbon Abundances in

Chondritic Meteorites

Abstract. Combustion analyses of total carbon in chondritic meteorites indicate a fractionation of this element between the various types of chondrites. The median values for the percentage of carbon by weight are 0.40 for enstatite chondrites, 0.09 for olivine-bronzite chondrites, and 0.08 for olivine-hypersthene chondrites. Olivine-pigeonite chondrites show great variations in their carbon contents.

The abundances of carbon have been determined in 18 chondritic meteorites by a combustion technique. Powdered samples ranging from 0.25 to I g in weight were burned in an oxygen at- mosphere in a LECO induction furnace. One gram of iron chips and 1 g of copper chips were added to acceler- ate the reaction. The combustion prod- ucts were carried in the oxygen stream through a MnO2 trap to remove any SO2 formed, and through a catalyst furnace to convert any carbon monox- ide to carbon dioxide. The total CO2

produced was determined by a modified Orsat technique (see 1) in a LECO model 572-100 carbon analyzer, by dis- solving the CO2 from the oxygen car- rier in a caustic solution and measuring the volume of CO, lost.

The technique gives reproducible and

16 JULY 1965

rier in a caustic solution and measuring the volume of CO, lost.

The technique gives reproducible and

16 JULY 1965

accurate results, as indicated by the results on two Thorn Smith steel stand- ards. Four analyses of one standard

ranged from 0.41 to 0.43 percent carbon by weight, as compared with the standard value of 0.40, and four analyses of the second standard ranged from 0.50 to 0.51, as compared with the standard value of 0.49.

The data obtained from the abundances of carbon in enstatite, olivine- bronzite, olivine-hypersthene, and carbonaceous and noncarbonaceous olivine-pigeonite chondrites are given in Table 1. Replicate analyses on the same batch of powdered meteorite for selected samples provide an index of the reproducibility of the method.

Previous determinations of carbon in chondrites have been largely limited to the three classes of carbonaceous chondrites and the enstatite chondrites. The presence of carbon in most of these meteorites is readily apparent and hence conducive to analysis. Carbon- aceous chondrites of the rarer types were not analyzed in this investigation because of their rarity and the ap- parently reasonable values obtained in previous analyses. Values obtained by other investigators (2-5) for the same meteorites analyzed in this investigation are indicated in Table 1.

As indicated by Table 1 and Fig. 1, in which carbon abundances reported here and abundances for carbonaceous chondrites reported by Boato (3) and Wiik (4, 5) are plotted, individual meteorites within a given classification generally show a broad range of values. This type of distribution makes a calculated mean for a meteorite group of questionable value for comparison purposes; each individual meteorite should be considered separately, and representative specimens for each class should be used in comparisons. For the specimens used in this investigation, the median values for the carbon content were: enstatite chondrites, 0.40; olivine-bronzite chondrites, 0.09; olivine- bronzite chondrites, 0.09; olivine-hypersthene chondrites, 0.08. These values are close to those selected by Craig (6) for the average carbon content of the olivine-bronzite (high-iron group) chondrites of 0.07 and the olivine-hypersthene (low-iron group) chondrites of 0.03 and the average value of 0.29 percent carbon selected by Wood (7) for

accurate results, as indicated by the results on two Thorn Smith steel stand- ards. Four analyses of one standard

ranged from 0.41 to 0.43 percent carbon by weight, as compared with the standard value of 0.40, and four analyses of the second standard ranged from 0.50 to 0.51, as compared with the standard value of 0.49.

The data obtained from the abundances of carbon in enstatite, olivine- bronzite, olivine-hypersthene, and carbonaceous and noncarbonaceous olivine-pigeonite chondrites are given in Table 1. Replicate analyses on the same batch of powdered meteorite for selected samples provide an index of the reproducibility of the method.

Previous determinations of carbon in chondrites have been largely limited to the three classes of carbonaceous chondrites and the enstatite chondrites. The presence of carbon in most of these meteorites is readily apparent and hence conducive to analysis. Carbon- aceous chondrites of the rarer types were not analyzed in this investigation because of their rarity and the ap- parently reasonable values obtained in previous analyses. Values obtained by other investigators (2-5) for the same meteorites analyzed in this investigation are indicated in Table 1.

As indicated by Table 1 and Fig. 1, in which carbon abundances reported here and abundances for carbonaceous chondrites reported by Boato (3) and Wiik (4, 5) are plotted, individual meteorites within a given classification generally show a broad range of values. This type of distribution makes a calculated mean for a meteorite group of questionable value for comparison purposes; each individual meteorite should be considered separately, and representative specimens for each class should be used in comparisons. For the specimens used in this investigation, the median values for the carbon content were: enstatite chondrites, 0.40; olivine-bronzite chondrites, 0.09; olivine- bronzite chondrites, 0.09; olivine-hypersthene chondrites, 0.08. These values are close to those selected by Craig (6) for the average carbon content of the olivine-bronzite (high-iron group) chondrites of 0.07 and the olivine-hypersthene (low-iron group) chondrites of 0.03 and the average value of 0.29 percent carbon selected by Wood (7) for the enstatite chondrites. It is interesting to note the parallelism of the carbon abundances in the chondritic meteorites

the enstatite chondrites. It is interesting to note the parallelism of the carbon abundances in the chondritic meteorites

- --UW ENSTATITE CHONDRITES

- -- - -- OLIVINE BRONZITE CHONDRITES

t I I I -: I OLIVINE HYPERSTHENE CHONDRITES

t I.--- . i--- OLIVINE PIGEONITE CHONDRITES

OLVYINEPIGEONITE CHONDRITES i - -' ; (CARBONACEOUS)

CARBONACEOUS CHONDRITES (TYPE )II 11 11 11 ---

CARBONACEOUS CHONDRITES (TYPE I ) I ---

- --UW ENSTATITE CHONDRITES

- -- - -- OLIVINE BRONZITE CHONDRITES

t I I I -: I OLIVINE HYPERSTHENE CHONDRITES

t I.--- . i--- OLIVINE PIGEONITE CHONDRITES

OLVYINEPIGEONITE CHONDRITES i - -' ; (CARBONACEOUS)

CARBONACEOUS CHONDRITES (TYPE )II 11 11 11 ---

CARBONACEOUS CHONDRITES (TYPE I ) I ---

Fig. 1. Carbon distribution in chondrites (percentage by weight). Fig. 1. Carbon distribution in chondrites (percentage by weight).

Table 1. Abundances of carbon in meteorites.

Previous Abundance Meteorite (% by wt.) results (% by wt.) (% by wt.)

Table 1. Abundances of carbon in meteorites.

Previous Abundance Meteorite (% by wt.) results (% by wt.) (% by wt.)

Hvittis Hvittis Enstatite chondrites

0.25 Enstatite chondrites

0.25 Khairpur .31 0.41 (5) Abee .40 .38 (2) Indarch .41 .4 (3)

.31 (4)

.43 (4) Atlanta .43 .50 (5)

Olivine-bronze chondrites Allegan .014

.018 Richardton .040 0.02 (3)

.040

.046

.049

.052

.052 Forest City .072 0.08 (3)

.072 Saline .11

.12 Kesen .15

.15

.16

.16 Beardsley .19

Olivine-hypersthene chondrites Bruderheim .040 New Concord .040

.046 Leedey .054 Dhurmsala .072

.085 Marion, Iowa .080

.085 Farmington .11

.11 Ergheo .11

.12 Modoc .18

.18 Olivine-pigeonite chondrites

Karoonda .10 Chainpur .57 Mokoia (carbo-

naceous) .75 0.84 (3)

Khairpur .31 0.41 (5) Abee .40 .38 (2) Indarch .41 .4 (3)

.31 (4)

.43 (4) Atlanta .43 .50 (5)

Olivine-bronze chondrites Allegan .014

.018 Richardton .040 0.02 (3)

.040

.046

.049

.052

.052 Forest City .072 0.08 (3)

.072 Saline .11

.12 Kesen .15

.15

.16

.16 Beardsley .19

Olivine-hypersthene chondrites Bruderheim .040 New Concord .040

.046 Leedey .054 Dhurmsala .072

.085 Marion, Iowa .080

.085 Farmington .11

.11 Ergheo .11

.12 Modoc .18

.18 Olivine-pigeonite chondrites

Karoonda .10 Chainpur .57 Mokoia (carbo-

naceous) .75 0.84 (3) .65 (3) .47 (4) .65 (3) .47 (4)

with the abundances of Hg, Tl, Pb, and Bi as noted by Reed et al. (8). These volatile elements are relatively depleted in the ordinary olivine-bronzite and olivine-hypersthene chondrites as compared to the enstatite, olivine-pigeonite, and carbonaceous chondrites.

CARLETON B. MOORE CHARLES LEWIS

Arizona State University, Tempe

317

with the abundances of Hg, Tl, Pb, and Bi as noted by Reed et al. (8). These volatile elements are relatively depleted in the ordinary olivine-bronzite and olivine-hypersthene chondrites as compared to the enstatite, olivine-pigeonite, and carbonaceous chondrites.

CARLETON B. MOORE CHARLES LEWIS

Arizona State University, Tempe

317

I I

Documents

Helium-Uranium Ratios for Pleistocene and Tertiary Fossil Aragonites